The invention relates to fuel flowmeters and provides accurate and reliable flow measurement of #2 diesel fuel.
Typical systems for net flow applications use two flowmeters, two temperature compensated sensors, and an indicator or datalogger to provide net rate and total. However, prior art flowmeters stop functioning when debris accumulates in the fuel supply lines thereby disrupting continuous fuel flow.
The present invention solves the prior art problems by maintaining continuous flow even if the flowmeter is jammed by debris in the fuel supplyxe2x80x94the bypassed fuel flow is sufficient to maintain engine operation from idle to full load.
The new system provides accurate and reliable flow measurement of #2 diesel fuel. For net flow applications, the typical system uses two flowmeters, two temperature compensated sensors, and an indicator or datalogger to provide net rate and total. The four available sizes are consistent with the ranges needed to provide flow measurement for most large diesel engines. Each meter is designed and built specifically for this application to provide optimal results. Additionally, its specialized internal geometry allows fuel flow to continue, even if the flowmeter is jammed by debris in the fuel supplyxe2x80x94the bypassed fuel flow is sufficient to maintain engine operation from idle to full load.
The new systems connects to xc2xcxe2x80x3, xc2xdxe2x80x3 and 1xe2x80x3 line sizes specifically for #2 diesel fuel. Reference accuracy is xc2x11% of net fuel consumption from 40xc2x0 F. to 160xc2x0 F. (4xc2x0 C. to 70xc2x0 C.) operating temperature for the two meter system (supply/return). The meters have built-in inlet flow conditioners for improved accuracy and piping insensitivity. The new system provides fuel temperature measurement and a temperature compensated output to 60xc2x0 F. Each meter has only two moving parts and provides in-line maintenance and easy installation. Non-clogging meters prevent starving the engine if a flowmeter jams. The system is a non-intrusive sensor and provides pulse-out or serial communications.
To make the system more accurate, software is imbedded in transmitters connected to the meters.
During calibration, each meter is subjected to a series of actual flow conditions at many flow rates and temperatures spanning the entire range for which the flowmeter is being calibrated. A typical calibration may consist of 84 data points.
Each data point consists of a temperature, a raw flow sensor signal (typically a square wave frequency representing the rotational velocity of the flowmeter impellers), and a xe2x80x9cmaster meterxe2x80x9d flow rate. The xe2x80x9cmaster meterxe2x80x9d flow rate represents the actual volumetric flow rate, compensated for temperature, viscosity and density effects such that the volumetric rate given is what it would have been had the temperature been 60xc2x0 F.
The xe2x80x9crawxe2x80x9d calibration data can be considered to form a three-dimensional surface, with raw flow sensor frequency and temperature representing the x and y coordinates, and the xe2x80x9cmaster meterxe2x80x9d flow rate representing the z coordinate (in a three-dimensional Cartesian coordinate system).
The data is then extended through extrapolation to make the surface (its projection onto the x-y plane) appear rectangular when viewed along the z axis.
This curved surface, which represents the actual flow rate as a function of frequency and temperature, is divided into a number of subsections. Each subsection is then xe2x80x9cfitxe2x80x9d by a multi-variate polynomial of sufficient order, whose coefficients are determined by standard mathematical means. Additional constraints are placed on the polynomials such that the ends which join with adjacent sections must match in value, first and second derivatives. This insures that the surface being modeled by the series of polynomials has no discontinuities at the transition points from one sub-section of the surface to the next.
This method of representing a continuous data set by a series of polynomials is known as splining, and the individual subsections are known as splines.
Knowing the frequency input from the flowmeter sensor and the temperature of the substance being measured, one can select the set of coefficients of the multi-variate polynomial which represent the subsection of the surface of interest. The frequency and temperature are applied to the polynomial using the appropriate coefficients, and the polynomial is evaluated, resulting in a close reproduction of the flow rate experienced by the flowmeter under the same conditions during calibration.
The order of the polynomials, the number of splines and the placement of the xe2x80x9cbreakpointsxe2x80x9d between one spline and the next can all be varied to optimize the xe2x80x9cfitxe2x80x9d of the splined model to more closely reproduce the resulting data obtained during calibration.
These and further and other objects and features of the invention are apparent in the disclosure, which includes the above and ongoing written specification, with the drawings.